LangGraph Q&A with Neo4j Graph Database

A showcase implementation demonstrating how to build an intelligent Q&A system using LangGraph and Neo4j Graph Database. This project combines the power of Large Language Models (LLMs) with graph database capabilities to answer questions about movie data.

🌟 Features

• Intelligent Movie Database Q&A: Ask natural language questions about movies, actors, directors, and genres
• Graph-based Knowledge Representation: Leverages Neo4j's graph capabilities for complex relationship queries
• Multi-stage Processing Pipeline: Uses LangGraph for dynamic orchestration of the Q&A process
• Query Validation and Self-correction: Intelligent validation and correction of generated Cypher queries
• Semantic Similarity: Uses example-based learning for improved query generation
• Smart Guardrails: Ensures the system only answers movie-related questions

📋 Requirements

• Python 3.8+
• Neo4j Database (accessible via connection string)
• OpenAI API Key
• LangSmith API Key (for tracing)

🔧 Installation

1. Clone the repository:

   git clone https://github.com/yourusername/langgraph_qa_with_graph_db_showscase.git
   cd langgraph_qa_with_graph_db_showscase

2. Install dependencies:

   pip install -r requirements.txt

3. Set up environment variables:

   # Create a .env file with the following variables
   OPENAI_API_KEY=your_openai_api_key
   LANGSMITH_API_KEY=your_langsmith_api_key
   LANGSMITH_TRACING=true
   LANGSMITH_PROJECT=GRAPH-QA
   NEO4J_URI=neo4j://localhost:7687  # Adjust as needed
   NEO4J_USERNAME=neo4j
   NEO4J_PASSWORD=your_password

🚀 Usage

Run the application:

python graph_qa.py

Example Questions

You can ask questions like:

• "Which actors played in the movie Casino?"
• "How many movies has Tom Hanks acted in?"
• "List all the genres of the movie Schindler's List"
• "Which actors have worked in movies from both the comedy and action genres?"

🧠 How It Works

This system follows a multi-stage process to answer questions:

1. Guardrails: Determines if the question is movie-related
2. Cypher Generation: Transforms the natural language question into a Cypher query
3. Validation: Checks the Cypher query for errors
4. Correction: Fixes any identified errors in the query
5. Execution: Runs the query against the Neo4j database
6. Answer Generation: Creates a natural language answer based on database results

Architecture Diagram

Input Question → Guardrails → Generate Cypher → Validate → Correct → Execute → Generate Answer
                     ↓             ↑               ↓         ↑
                    End         Correction ←── Validation

💾 Database Schema

The movie database includes:

• Movie nodes: With properties like title, released date, and IMDB rating
• Person nodes: Representing actors and directors
• Genre nodes: Different movie genres
• Relationships: ACTED_IN, DIRECTED, IN_GENRE

Graph Visualization

Developed by extrawest. Software development company